written in July 2000, updated in February 2004
Since the recent U.S. Supreme Court ruling (California Democratic Party vs. Jones) that found that a state cannot mandate that political parties use a blanket primary to choose its candidates, some blanket primary supporters have been proposing that their states use some form of a Louisiana-style nonpartisan primary. The following report analyzes some of the electoral results and impacts of Louisiana's nonpartisan primary. Its findings demonstrate that Louisiana's system limits voter choices and can elect unrepresentative candidates. These results should give pause to those who see a Louisiana-style nonpartisan primary as a suitable replacement for a blanket primary. They instead should consider other alternatives such as instant runoff voting and full representation.
The 2000 U.S. Supreme Court decision, California Democratic Party v. Jones, ruled that a state cannot mandate that political parties use a blanket primary to choose its candidates. The blanket primary states of California, Washington and Alaska ultimately had to change their primary systems, and other states growing interested in blanket primaries have turned to look to other reform approaches.
As is so often the case in matters of jurisprudence, the Court was trying to balance competing demands. On the one hand, there is the right of political parties to guarantee the integrity of their candidate-selection process for their registered voters. But on the other hand there are the insistent voices of millions of voters, many of them independents (the fastest growing demographic of voters), saying they're tired of poor choices in general elections -- sometimes due to non-competitive races that we term "monopoly politics," sometimes due to the two major party nominees poorly representing the views of voters in the political center. These voters are demanding better electoral choices -- or simply not showing up on election day.
"In an era of dramatically declining voter participation, states should be free to experiment with reforms designed to make the democratic process more robust by involving the entire electorate in the process of selecting those who will serve as government."
Justice John Paul Stevens, dissent in Jones (2000)
Proponents of blanket primaries tend to cite at least two reasons for their support for this reform:
1. Moderation: Blanket primaries act as a moderating influence on the candidate selection process, with non-party voters in the primary boosting candidates from more moderate parts of the spectrum of their respective political parties; and
2. Choices: Blanket primaries tend to give voters more choice at the ballot box, since voters can cross over from party to party and from race to race.
Some political observers believe that the U.S. Supreme Court in the Jones case has indicated that some version of a Louisiana-type of nonpartisan blanket primary may be constitutionally acceptable. In Louisiana, voters can choose any candidate from any political party in the opening round of voting. If one candidate wins a majority of votes cast, that candidate is declared the winner. If no candidate wins a majority, the top two vote-getters -- regardless of party affiliation -- move on to the general election. Presumably, blanket primary supporters now pushing for a Louisiana-style nonpartisan primary hope that the latter will provide at least some of the benefits of a blanket primary.
However, a survey of election results in Louisiana reveals that Louisiana's nonpartisan primary may actually work contrary to both of the above-stated goals of blanket primary supporters.
A Closer Look at Results in Louisiana:
The Louisiana primary system does not have a moderating influence. Louisiana's nonpartisan primary often appears to have had the opposite effect of the goal of electing more moderate winners. In fact, Louisiana is notorious for candidates like David Duke, the former leader of the Ku Klux Klan, winning or coming close to winning high public office. A Republican state legislator, Duke ran a strong second in the 1990 U.S. Senate election and gained a spot in the runoff election in the governor's race in 1991. In that 1991 runoff, he faced Edwin Edwards, a former governor with a history of suspected corruption. Indicating the polarized nature of the choice between Duke and Edwards, a popular bumper sticker in favor of Edwards was: "Vote the Crook: It's Important."
In the 1995 governor's race, sixteen candidates ran in the opening round, including four major candidates who ultimately won at least 18% of the vote. The two most ideologically extreme major candidates were Mike Foster, a conservative Republican who earned Pat Buchanan's endorsement and inherited much of David Duke's constituency, and Cleo Fields. a leading liberal Democrat in the Congressional Black Caucus. They advanced to the runoff election with a combined vote of only 45% of votes casts, with the more centrist vote split among other candidates. Foster ultimately was elected in the runoff election.
A Louisiana-style nonpartisan primary easily can produce these kind of results because in a large field of candidates, the top two vote-getters can have relatively few votes. In a multi-candidate field, this rule tends to favor non-moderate candidates with the strongest core support that can be narrow rather than broad. This lack of moderation is the exact opposite of one of the goals of blanket primary proponents like former California Congressman Tom Campbell.
The Louisiana primary system does not give voters more choice. Louisiana's nonpartisan primary can reduce voters' choices at the ballot box rather than increase them. Most importantly, few races have gone to a second round of voting, meaning that, until the state law was changed in 1998 to hold the primary in November and the general election in December, almost all federal races were decided in October in the opening primary round of election rather than in the general election in November. In fact, most races were won without any competition whatsoever. In Louisiana's congressional elections in 1998, for example, incumbents faced no opponents in five out of seven U.S. House seats and didn't even appear on the ballot . A sixth incumbent easily defeated two candidates from his party in the opening round. The final incumbent faced one challenger, whom he narrowly defeated. A total of 10 candidates ran for seven seats, with only one remotely close race.
In 1996, three out of seven House seats were uncontested, and two more were won by "landslide" victory margins of more than 20%. In 1994, there were no general elections for congressional seats because all of the races were decided in the opening primary round -- in these races six out of seven House races were won by landslides with an average victory margin of 86 percent. In 1992, six out of seven House races were won by landslides with huge victory margins. Louisiana has been using its nonpartisan primary since 1977; in the eleven election years since that time, only a single congressional incumbent has been defeated.
Perhaps it should be no surprise that throughout the 1990's, Louisiana often ranked last in the nation in voter turnout in U.S. House election. In 1998, only 2 out of 7 races were contested. The turnout in these races was typical of House races, but more problematically, voters in most districts didn't even have their representatives appear on the ballot. This system certainly hasn't given voters better choices, and it certainly hasn't encouraged voter participation. In addition, Louisiana's election and redistricting methods have produced startling disproportions between the percentage of seats won compared to the percentage of popular votes won.
The Louisiana system produces other perverse impacts. For instance, in 1996 the 7th US House district opening primary resulted in two Democrats finishing highest -- with a combined vote of only 48% of votes cast -- and facing off in the general election. No Republican or third party candidates thus appeared in the general election. In 1992, the 4th congressional district primary resulted in two Democrats reaching the runoff election (without Republican or third party candidate appearing on the November ballot), while the 6th congressional district resulted in two Republicans reaching the runoff election (without Democrats or third party candidates on the November ballot).
As a general rule, since only the top two finishers in the opening primary have had any chance to advance to a runoff election, it proved to be very difficult for minor party candidates to appear on Louisiana's November election ballot for federal and state elections. Not surprisingly, there have been no effective minor parties in Louisiana in the years since the present election system was put into place in 1978. There have been only a handful of minor party candidates on the ballot with a party label for any federal office (except President) in that period.
Finally, Louisiana is likely to face new problems with its change in 1998 to have the opening primary round at the time of the November election. This change certainly will ensure higher voter turnout in the opening round -- important, given the fact that most elections have been decided in that round -- but will also result in any runoff elections taking place in December, when turnout is often going to be much lower. For example, in 1992, a U.S. Senate race in Georgia went to a December runoff, and turnout dropped by more than half between November (which took place at the time of the presidential election) and the decisive runoff.
Thus, from the point of view of voters, nonpartisan primaries in Louisiana have severely limited the range of choices available in the general elections. When given a choice, voters typically are limited to two candidates, who frequently are not moderates and sometimes are actually from the same political party..
Better solutions. Fortunately, there are better ways to produce the goals of blanket primaries. One approach would be to use instant runoff voting. For legislative elections, another approach would be full representation systems-- like those touted by blanket primary advocate Tom Campbell of California.
Instant runoff voting:
There are two ways that instant runoff voting could meet the goals of blanket primary advocates: in primaries or in the general election. Note that the instant runoff produces a majority winner in one round of voting -- as opposed to two-round runoffs. Two-round runoffs suffer the defect of requiring voters to trek to the polls a second time, and at a certain point voter fatigue sets in. Also, candidates must raise money for two elections, negating one of the goals of campaign finance reform, and taxpayers must fund two rounds of election. With instant runoffs, voters rank candidates to indicate their first choice as well as their runoff choices ahead of time. Ballots are counted like a series of runoffs, and the final result is a winner who has the support of a majority of voters. London uses the instant runoff to elect its mayor, and Australia and the Republic of Ireland have used it for decades in their national elections. Two American localities passed instant runoff voting ballot measures in 1998 and 1999.
Using the instant runoff voting in primaries would ensure that party nominees were not opposed by a majority of primary voters. In big candidate fields - as often happens when a seat is open, and one party is heavily favored to take the seat in the general election -- it would promote coalition-building. The end result would be stronger, more representative nominees.
The instant runoff also could be used in the general election in such a way that primaries could be eliminated entirely. Parties could develop their own mechanism of nominating candidates -- perhaps at a convention or a privately-financed primary -- but the real choice would be decided in the general election, when all voters participate. A major party candidate who did not win his or her party's nomination could still seek the office in the general election because instant runoff voting can produce a majority winner no matter how many candidates run. This option would be attractive to those who would like to see shorter campaigns and have the decisive election occur in the election with maximum participation.
In legislative elections, "winner take all" rules could be modified. For example, Illinois elected its state house of representatives by cumulative voting from 1870 to 1980. Candidates ran in "multi-seat" districts with three representatives each -- meaning that there were a fewer number of larger districts. Voters had three votes and could allocate those votes however they wished; if they really liked one candidate, they could give all their votes to that one candidate. The impact of cumulative voting was that if about a quarter of voters in a district supported one party, they would have the votes to elect one of three seats. The party with a majority of votes would be able to win a majority of two seats, but could win all three seats only if it had an overwhelming majority that is rare in any one part of a state.
In Illinois, cumulative voting led to the major parties typically accepting that they did not have enough votes to win all the seats in a given district. The parties thus generally would nominate only two candidates -- using cumulative voting in the primary as well as the general. Voters in the general election then were typically presented with four major candidates who reflected somewhat different parts of the spectrum. The result was fuller representation of the electorate and better choices for voters, both in the primary and general election. Among those now supporting return of cumulative voting in Illinois are the Chicago Tribune, former governor Jim Edgar, former congressman and federal judge Abner Mikva and leading current elected officials in both parties. Leading blanket primary advocate Tom Campbell, a former Republican congressman from California, testified in 1999 in favor of Rep. Mel Watt's State's Choice of Voting Systems Act (a bill to allow states to use full representation systems in congressional elections), based largely on his positive experience with cumulative voting in Illinois.
There are other full representation systems that would create better opportunities for small parties to win seats and give voters an even wider range of choice.
For better or worse, the Supreme Court has declared it unconstitutional for states to mandate that parties use blanket primaries. Advocates of blanket primaries should resist any temptation toward adopting a version of Louisiana's flawed nonpartisan primary system and instead consider other approaches like instant runoff voting and full representation systems. They would provide better choices to voters and fuller representation of the spectrum of opinion.
Read more analysis of blanket primaries:
Blanket primary has other solutions, and it's not Louisiana's
Commentary by Steven Hill in Roll Call, July 25, 2000 (read)
A win/win alternative to the blanket primary
Letter to editor by Janet Anderson in the Seattle Post-Intelligencer, July 22, 2000 (read)